Isoquinolinone synthesis by SNAr reaction: A versatile route to imidazo[4,5-h]isoquinolin-9-ones

Article abstract of DOI:10.1016/S0040-4039(02)01802-6

Reaction of 2-chlorobenzonitriles with β-ketoesters in an S_NAr reaction, followed by cyclization in acid provides a versatile route to isoquinolones. Starting from 2,6-dichloro-3-nitrobenzonitrile 7, sequential displacement of the chlorines by

Full text of DOI:10.1016/S0040-4039(02)01802-6

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 7553–7556
Isoquinolinone synthesis by SNAr reaction: a versatile route to
imidazo[4,5-h]isoquinolin-9-ones
Roger J. Snow,a,* Tanja Butz,a Abdelhakim Hammach,a Suresh Kapadia,b Tina M. Morwick,a
Anthony S. Prokopowicz, III,a Hidenori Takahashi,a Jonathan D. Tan,b Matt A. Tschantza and
Xiao-Jun Wangb
aDepartment of Medicinal Chemistry, Boehringer Ingelheim Pharmaceuticals Inc., 900 Ridgebury Road, Ridgefield, CT 06877,
USA
bChemical Development, Boehringer Ingelheim Pharmaceuticals Inc., 900 Ridgebury Road, Ridgefield, CT 06877, USA
Received 3 June 2002; revised 16 August 2002; accepted 23 August 2002
Abstract—Reaction of 2-chlorobenzonitriles with b-ketoesters in an SNAr reaction, followed by cyclization in acid provides a
versatile route to isoquinolones. Starting from 2,6-dichloro-3-nitrobenzonitrile 7, sequential displacement of the chlorines by an
amine and a b-ketoester leads to imidazo[4,5-h]isoquinolin-9-ones 1, a new class of kinase inhibitor. © 2002 Elsevier Science Ltd.
All rights reserved.
Recently we disclosed a series of phenylamino imi-
dazo[4,5-h]isoquinolin-9-ones,1 exemplified by 1, as a
new class of potent inhibitors of the tyrosine kinase
p56lck (lck). These compounds are potentially useful
therapeutic agents for treating autoimmune diseases.
The discovery of 1 resulted from a structure–activity
relationship and molecular modeling study of a screen-
ing lead, 2.
tion of the C7 carbonyl, followed by acid treatment to
induce a Wagner–Meerwein rearrangement. This had
two severe limitations. Firstly, the synthesis of 2
required 10 steps from commercial starting materials.
Secondly, the rearrangement to produce 1 was success-
ful only in the dimethyl case. Our analysis of the
interaction of 1 with lck1 indicated that modification of
the substituents in the pyridone ring should be benefi-
cial, so we designed a synthetic route that allowed wide
variation at these positions, and required fewer steps
after introduction of this diversity element. This
required a flexible approach to the isoquinolone moiety
of 1, and efficient access to the highly substituted
central benzene ring.
Our approach to the isoquinolone (Scheme 1) was
based on the SNAr reaction of 2-chlorobenzonitrile 3
with b-ketoester 4a to give 5, followed by cyclization
under acid conditions. The ready availability of b-
ketoesters would allow a wide range of substitution at
Following this discovery, we needed an improved syn-
thesis of 1 and its analogs for further studies. The
original route to 1 proceeded from 2 by selective reduc-
Scheme 1. Approaches to isoquinolones.
* Corresponding author. Tel.: +1-203-798-4060; fax: +1-203-791-6072; e-mail: rsnow@rdg.boehringer-ingelheim.com
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)01802-6

7554
R. J. Snow et al. / Tetrahedron Letters 43 (2002) 7553–7556
both the 3 and 4 positions of the isoquinolone, and
would offer much greater versatility than isoquinolone
syntheses based on alkynes2 or directed lithiation.3 A
conceptually similar route to isoquinolones by a photo-
chemical reaction of 2-bromobenzamide and enolates of
simple ketones has been reported.4 Sommer5 has
described SNAr reactions of 2-chlorobenzonitriles,
including 3, with phenylacetonitriles under ionic condi-
tions. We expected that the nitro group in 3 would
enhance its reactivity, and also provide one of the
benzimidazole nitrogens. As a trial reaction, 3 was
reacted with 4a in the presence of K2CO3 in DMF at
room temperature to give the desired adduct 5 in 47%
yield. Heating 5 in a H2SO4/water/acetic acid (1:1:3)
mixture at 100°C brought about hydrolysis, decarboxyl-
ation and cyclization in one step to provide 6 in quanti-
tative yield.
kind of system.6 Treatment of 7 with ammonia in
ethanol in a sealed tube or methylamine in THF led to
the formation of 8 (68%) and 9 (83%), respectively. In
both cases the product crystallized from the reaction
mixture in >95% purity in the indicated yield. The
mother liquor from 9 contained more product, along
with the regioisomer (4–6% yield) and a small amount
of the diamine (2%). The position of the amine was
confirmed chemically by conversion to the benzimi-
dazole.
With 8 and 9 in hand we investigated the SNAr reaction
with ketoester 4a (Table 1). Using K2CO3 in DMF, the
products 10a and 11a were formed in 46 and 56% yield,
respectively. The optimum conditions used 1.1 equiv. of
base and 1.5 equiv. of ketoester. Functionalized
ketoesters 4c and 4g gave similar yields. Subsequently it
was found that the reaction was faster and generally
higher yielding with Cs2CO3. Potassium t-butoxide also
worked well, giving 11a in 69% yield. Polar solvents
such as DMF or DMSO were preferred. Amine bases
or phase transfer conditions were unsatisfactory. In
contrast to the other examples, reaction of ethyl substi-
tuted ketoester 4b with 8 gave poor yields of 10b with
either K2CO3 or Cs2CO3. In this case a by-product that
appeared to arise from dimerization of 8 was observed.
However, a much better yield of the N-methyl analog
11b was obtained when 4b was reacted with 9 in the
presence of Cs2CO3. In other cases, the main by-prod-
Having found a route to the isoquinolone, the next
challenge was introduction of the benzimidazole nitro-
gens of 1. The synthesis of 2 took six steps to insert
these nitrogens. An attractive commercial precursor
that contains the 1,2,3,4-tetrasubstitued benzene
required for the central ring is 2,6-dichloro-3-nitro ben-
zonitrile (7). We reasoned that it might be possible to
displace the two chlorines sequentially, using one to
introduce the second nitrogen, and the other to carry
out the SNAr reaction (Scheme 2). There is evidence
that the chlorine at C2 is more readily displaced in this
Scheme 2. Synthesis of tetrasubstituted benzene precursor.
Table 1. Examples of SNAr reaction
Ketone
Nitrile
R2
R3
R4
Conditionsa
Product
Yieldb
4a
4a
4a
4a
4b
4b
4b
4c
4c
4d
4e
4f
8
9
9
9
8
8
9
8
9
9
9
8
9
8
8
9
9
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
CO2Et
H
CO2tBu
CO2tBu
CO2tBu
Ph-4-OMe
CO2Et
H
Me
Me
Me
Me
Et
Et
Et
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
CH2OBn
Me
A
A
B
C
A
B
B
A
B
B
A
A
B
A
A
B
B
10a
11a
11a
11a
10b
10b
11b
10c
11c
11d
11e
10fc
11fc
10g
10h
11i
46
56
80
69
25
16
59
40
70
55
84
49
58
46
41
49
85
CH2CO2Et
CH2CO2Et
(CH2)2CO2Et
CO2Me
H
H
H
H
4f
4g
4h
4i
(CH2)3
C(ꢀO)(CH2)3
4j
11j
a Conditions: A: K2CO3, DMF, rt 24–48 h. B: Cs2CO3, DMF, rt, 2–6 h. C: KOtBu, DMSO, 25–40°C.
b Yields of pure product after chromatography.
c R2=H, isolated after cleavage of CO2tBu by treatment with TFA.

R. J. Snow et al. / Tetrahedron Letters 43 (2002) 7553–7556
7555
Scheme 3. Conversion of ketoesters to imidazo[4,5-h]isoquinolin-9-ones.
Table 2. Yields for the conversion of ketoesters to imidazo[4,5-h]isoquinolin-9-ones
SM
R1
R2
R3
R4
Benzimidazole
Yielda
Product
Yield
10a
11a
11c
11d
11e
10f
11f
11i
H
CO2Et
CO2Et
CO2Et
CO2Et
H
H
H
CO2Et
Me
Me
Me
Me
Me
Me
Me
Me
Me
12a
13a
13c
13d
13e
12f
13f
13i
47
51
69
74
77
87
61
92
1a
72
88
Me
Me
Me
Me
H
14a
14e
14d
14e
1f
CH2CO2Et
(CH2)2CO2Et
CO2Me
H
H
78b
62b
70c
48
Me
Me
14f
14i
90
43
(CH2)3
a Overall isolated yield for conversion of 10 to 12 or 11 to 13, respectively.
b Yield after re-esterification with EtOH/H2SO4.
c Cyclization in H2SO4, 0°C–rt.
This approach represents a significant improvement in
the synthesis of 1. The number of steps was reduced
from twelve to six, and the added flexibility provided
access to analogs for SAR studies, including com-
pounds with functional groups suitable for further elab-
oration. Certain analogs showed enhanced biological
activity, details of which will be reported elsewhere.7
ucts were dark baseline material and unreacted nitrile.
While most examples used ketoesters, diketones (4j)
and ketones with additional stabilization by an aro-
matic ring (4h) also gave satisfactory results. The suc-
cess of this reaction on a highly functionalized system,
even in cases where the product contains a quaternary
center, provided rapid access to the core of 1.
To convert 10a to 1, either the pyridone or the benzim-
idazole ring could be formed first. Closure of the pyri-
done ring, as in the model system, succeeded, but
subsequent steps were hampered by the extremely poor
solubility of the resulting isoquinolone. Several final
compounds were prepared in this way. A more satisfac-
tory procedure was to reduce 10 to the diamine and
convert this to the benzimidazole 12 before forming the
pyridone ring (Scheme 3). For 12 the diamine was
converted to the thiourea, then cyclized with DCC in a
separate step. For the N-methyl benzimidazoles 13,
treating the diamine with the isothiocyanate and HgO
in a one-pot procedure was found to be more reliable.
Closure of the pyridone ring by heating 12 and 13 in
acid gave 1 and 14 (Table 2). Side chain esters
hydrolyzed during the reaction, and were re-esterified
prior to isolation. The final product precipitated on
neutralization, in most cases in good yield, and did not
require further purification. One exception was ester
13e, which gave a mixture of the ester 14e and decar-
boxylated product 14f under the standard conditions.
To obtain 14e cleanly, 13e was treated with H2SO4
alone, without heating. To access 14f, the unsubstituted
ketone 11f, obtained from the t-butyl ester by treat-
ment with TFA, was treated under the standard condi-
tions. Attempts to hydrolyze the methyl ester of 13e
under basic conditions prior to cyclization led instead
to deacetylation.
Acknowledgements
We thank Drs. Neil Moss and Daniel Goldberg for
helpful discussions and critical reading of the
manuscript.
References
1. Snow, R. J.; Cardozo, M. G.; Morwick, T. M.; Busacca,
C. A.; Dong, Y.; Eckner, R. J.; Jacober, S.; Jakes, S.;
Kapadia, S.; Lukas, S.; Panzenbeck, M.; Peet, G. W.;
Peterson, J. D.; Prokopowicz, A. S., III; Sellati, R.; Tol-
bert, R. M.; Tschantz, M. A.; Moss, N. J. Med. Chem.
2002, 45, 3394–3405.
2. (a) Wu, M.-J.; Chang, L.-J.; Wei, L.-M.; Lin, C.-F. Tetra-
hedron 1999, 55, 13193–13200; (b) Sakamoto, T.; Miura,
N.; Kondo, Y.; Yamanaka, H. Chem. Pharm. Bull. 1986,
34, 2760–2765.
3. (a) Brun, E. M.; Gil, S.; Mestres, R.; Parra, M. Synthesis
2000, 273–280; (b) Fisher, L. E.; Muchowski, J. M.; Clark,
R. D. J. Org. Chem. 1992, 57, 2700–2705.
4. Bois-Choussy, M.; Beugelmans, R. Synthesis 1981, 729–
731.
5. Sommer, M. B.; Bergtrup, M.; Bogeso, P. J. Org. Chem.
1990, 55, 4817–4821.

7556
R. J. Snow et al. / Tetrahedron Letters 43 (2002) 7553–7556
6. Capps, D. B.; Dunbar, J.; Kesten, S. R.; Shillis, J.; Werbel,
L. M.; Plowman, J.; Ward, D. L. J. Med. Chem. 1992, 35,
4770–4778.
7. Goldberg, D. R.; Butz, T.; Cardozo, M. G.; Eckner, R. J.;
Hammach, A.; Huang, J.; Jakes, S.; Kapadia, S.; Kashem,
M.; Lukas, S.; Morwick, T. M.; Moss, N.; Panzenbeck,
M.; Patel, U.; Pav, S.; Peet, G. W.; Peterson, J. D.;
Prokopowicz, A. S., III; Snow, R. J.; Sellati, R.; Taka-
hashi, H.; Tan, J.; Tschantz, M. A.; Wang, X.; Wang, Y.;
Wolak, J.; Xiong, P. J. Med. Chem., submitted.